In this document, I have used the dataset called “USArrests”. This is a public dataset available in R. To see the dataset, use the library called “datasets”. For more details on the dataset see https://stat.ethz.ch/R-manual/R-devel/library/datasets/html/USArrests.html.

Use the following commands to install the dataset in your R-studio:

install.packages("datasets")
library(datasets)

Now, you can start using “USArrests” like any regular data.frame in R.

This is the arrest data from 1973. By this time, all 50 states had been part of US for about 15 years. The aim is to see if there is any dependency between when the state was acquired and the arrest history. Depending on your political stand point, you could argue that the arrest data is not the true measure of the crime commited, but for the purpose of this document we’ll assume that all arrests are valid and that most criminals were arrested.

The top 10 states with minimum arrests:

##       Total Arrests   Assault         Rape            Murder         
##  [1,] "North Dakota"  "North Dakota"  "North Dakota"  "North Dakota" 
##  [2,] "Vermont"       "Hawaii"        "Maine"         "Maine"        
##  [3,] "Wisconsin"     "Vermont"       "Rhode Island"  "New Hampshire"
##  [4,] "New Hampshire" "Wisconsin"     "West Virginia" "Iowa"         
##  [5,] "Iowa"          "Iowa"          "New Hampshire" "Vermont"      
##  [6,] "Hawaii"        "New Hampshire" "Wisconsin"     "Idaho"        
##  [7,] "Minnesota"     "Minnesota"     "Connecticut"   "Wisconsin"    
##  [8,] "Maine"         "West Virginia" "Vermont"       "Minnesota"    
##  [9,] "West Virginia" "Maine"         "Iowa"          "Utah"         
## [10,] "South Dakota"  "South Dakota"  "South Dakota"  "Connecticut"

The top 10 states with maximum arrests:

##       Total Arrests    Assault          Rape         Murder          
##  [1,] "Florida"        "North Carolina" "Nevada"     "Georgia"       
##  [2,] "North Carolina" "Florida"        "Alaska"     "Mississippi"   
##  [3,] "Maryland"       "Maryland"       "California" "Louisiana"     
##  [4,] "Arizona"        "Arizona"        "Colorado"   "Florida"       
##  [5,] "New Mexico"     "New Mexico"     "Michigan"   "South Carolina"
##  [6,] "California"     "South Carolina" "New Mexico" "Tennessee"     
##  [7,] "Alaska"         "California"     "Florida"    "Alabama"       
##  [8,] "South Carolina" "Alaska"         "Arizona"    "North Carolina"
##  [9,] "Nevada"         "Mississippi"    "Oregon"     "Texas"         
## [10,] "Michigan"       "Michigan"       "Missouri"   "Nevada"

There doesn’t seem to be much dependency on when the state was added to the country and the arrests made. For example, Hawaii was the last state to be added (August 21, 1959). But it has one of the lowest crimes. I can make the opposite argument for Arizona. It has one of the highest arrest records, and it was the third last state to be added (February 14, 1912). North Dakota was the 39th state (November 2, 1889) and this had the lowest crime in United States in 1973. The only conclusion that can be safely drawn is that there isn’t enough evidence to say that any crime can be explained by when the state was added to the country, and that by 1973, the effects, if any, had settled off.

Of course, it would be much more helpful to see exactly how many arrests were made a few years before the state was added, and a few years after. This would give a much clearer picture and which states felt the effects. But that’s an analysis for another day!

The statistically proficient readers will argue that this is not a scientific method to validate or reject a theory. While that is true, in this case, it is possible to draw the conclusion through sorting the dataset.